Google Screened

Open-Source Project Bolsters Self-Driving Advancements

Chester County Car Accident Lawyers discuss open-source software and how it can impact the self-driving vehicle movement. Open-source, which denotes software for which the original source code is made freely available to be redistributed and modified, is an approach taken by many software producers. Sometimes, proponents of this model simply feel that software codes should be made available to the public. Other times, open-source is driven by economic motivation. If software producers can develop a marketplace for their source codes, as the experts in that marketplace, they can make money by charging for their knowledge. They can provide advice about the software, as well as make money with add-ons.

Whatever the impetus for open-source, it has infiltrated the car industry, specifically, for driverless cars, also known as autonomous cars or self-driving cars. This technology allows for cars that are capable of sensing their environment and navigating without human input. Advocates of self-driving cars believe they might greatly reduce car accidents and fatalities.

SAE International (SAE) adopted a five-level definition of self-driving capabilities for cars. Level Zero is entirely driver operated with no autonomy, while Level Five would be able to autonomously negotiate complex environments like dirt roads. Currently, no self-driving capabilities have surpassed Level Four, which would remove the need for a driver, but still not be able to cover every possible condition, such as is being tested by Uber and others now.

Self-Driving Cars and Open-Source Programs

Three software companies, Openpilot, Uadacity, and Autoware, have attempted to open-source self-driving cars.

Openpilot provides an open-source hardware device called Comma Neo, which offers lane-keeping assist and adaptive cruise control functions. Based on the use of an Android phone, it has several sensory capabilities for a self-driving car, and should be installed where the rear-view mirror sits. The device can be built from directions which are shared at Comma.AI or Github. Openpilot’s creator admits that it cannot turn your car into a self-driving car, as the software’s capabilities are on par with SAE’s Level Two, which can assist the driver with cruise-control speed and lane-centering.

Udacity launched its self-driving car certificate program, Nanodegree, in an effort to educate engineers on the specifications for creating a self-driving car. They are aiming for a Level Four self-driving car. Udacity offers challenges and contests, allowing students to contribute source codes to the open-source project. These challenges provide great educational experiences for the students; however, it will take quite a bit of time to build a self-driving car with this tactic, as it depends on the interest of the student to contribute to this effort.

Autoware is being led by university researchers in Japan, using a ZMP Robo Car modeled after the Toyota Prius. With the direction of this research, it appears that Autoware is more focused on furthering self-driving algorithms, instead of producing a commercial product that can be sold in a marketplace. Autoware uses Velodyne LIDAR sensors, Point Grey Ladybug and Grasshopper cameras, and Javad RTK sensors for GPS information.

Self-Driving Cars and Overall Safety

While advancements with self-driving cars are apparent with these three open-source projects, all lack safety precautions. Most programmers lack the know-how to code for safety, and therefore neglect safety issues. Consequently, most individuals would have some reservations about being a passenger in a driverless car, especially after Uber’s fatal car crash in March 2018.

Self-driving taxi startup Voyage is aware of this major problem. Therefore, instead of focusing on open-source projects to create a self-driving car, its priority is to share its safety requirements, test scenarios, metrics, tools, and codes for its Level Four self-driving taxis.

Voyage’s initial release, Open Autonomous Safety (OAS), will be a GitHub storehouse with documents and code. OAS is complete with scenarios designed to stress-test cars in simulations and on streets.

The Voyage CEO admits that the initial OAS is far from perfect, but will allow programmers to further the advancement of safety requirements for self-driving cars.

Chester County Car Accident Lawyers at McCann Dillon Jaffe & Lamb, LLC Work Hard to Help You Obtain Justice

If you have been in a car accident as a result of a self-driving car, please contact a Chester County car accident lawyer at McCann Dillon Jaffe & Lamb, LLC for legal representation. Our Chester County car accident lawyers are dedicated to providing injured plaintiffs with skilled legal representation in personal injury and civil litigation matters. Contact us by calling 215-569-8488 or contact us online to schedule a free initial consultation. With offices across Philadelphia, Pennsylvania, we serve clients from the surrounding areas.